Keeping Variables Common To Frames

[Jared Greathouse]

Hey everyone. Consider the following two frames of data. The first is imported from my Google Drive, given to me by a colleague, the second from the Correlated of State Policy Database at IPPSR of Michigan State.

clear *


capture program drop gsheet
program define gsheet
    syntax anything , key(string) id(string)
    
    local url "https://docs.google.com/spreadsheets/d/`key'/export?gid=`id'&format=csv"
    
    copy "`url'" `anything', replace
    
    noi disp `"saved in `anything'"'
    
    import delim using `anything', clear
    
    erase `anything'
end

gsheet "`c(pwd)'\\test.csv" , key("12BT6vxYDf_00k6CjJPyuKY7LBs5L-ljVC8ORB6APXeQ") id("542334262")

frame rename default main

frame create fulldata


frame change fulldata


cls

import delim "https://ippsr.msu.edu/sites/default/files/cspp/correlates2-6.csv", clear varn(1) bindquote(strict)

keep if inrange(year,1970,2000)

What I want.... is to keep the variables in frame fulldata, on the condition that they are present in frame main. So, for example, total_expenditure_inc total_expenditure are both in each data frame, therefore, they both should stay in frame two and all the other ones be dropped. I didn't want to go through each predictor and check and see if it was there, I wanted to see if a more systematic way existed. How might I start this?

[Andrew Musau]

You can use extended macro functions:

frame change fulldata
frame main: qui ds
frame main: local main `r(varlist)'
qui ds
local fulldata `r(varlist)'
local drop: list main - fulldata
drop `drop'